A property-oriented design strategy for high performance copper alloys via machine learning

Wang, Changsheng; Fu, Huadong; Jiang, Lei; Xue, Dezhen

doi:10.1038/s41524-019-0227-7

Cited by 134 publications

(49 citation statements)

References 48 publications

(49 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One example in the recent applications in this area is a machine learning design system (MLDS) developed to search for copper alloys of high ultimate tensile strength (UTS) and electrical conductivity (EC) as target properties. [576] The results in this research have revealed the successful implementation of MLDS method in assisting inverse design of 8 new high-performance copper alloys with target UTS of 600 to 950 MPa and EC over 50 pct of International Annealed Copper Standards. A second example in this area, Wen et al [577] have developed a property oriented optimization strategy to search for large hardness as a desired property in Al-Co-Cr-Cu-Fe-Ni high entropy alloys (HEA).…”

Section: Machine Learning-assisted Alloy Microstructure and Propmentioning

confidence: 79%

Current Challenges and Opportunities in Microstructure-Related Properties of Advanced High-Strength Steels

Raabe

Sun

Silva

et al. 2020

Metall Mater Trans A

137

View full text Add to dashboard Cite

This is a viewpoint paper on recent progress in the understanding of the microstructure–property relations of advanced high-strength steels (AHSS). These alloys constitute a class of high-strength, formable steels that are designed mainly as sheet products for the transportation sector. AHSS have often very complex and hierarchical microstructures consisting of ferrite, austenite, bainite, or martensite matrix or of duplex or even multiphase mixtures of these constituents, sometimes enriched with precipitates. This complexity makes it challenging to establish reliable and mechanism-based microstructure–property relationships. A number of excellent studies already exist about the different types of AHSS (such as dual-phase steels, complex phase steels, transformation-induced plasticity steels, twinning-induced plasticity steels, bainitic steels, quenching and partitioning steels, press hardening steels, etc.) and several overviews appeared in which their engineering features related to mechanical properties and forming were discussed. This article reviews recent progress in the understanding of microstructures and alloy design in this field, placing particular attention on the deformation and strain hardening mechanisms of Mn-containing steels that utilize complex dislocation substructures, nanoscale precipitation patterns, deformation-driven transformation, and twinning effects. Recent developments on microalloyed nanoprecipitation hardened and press hardening steels are also reviewed. Besides providing a critical discussion of their microstructures and properties, vital features such as their resistance to hydrogen embrittlement and damage formation are also evaluated. We also present latest progress in advanced characterization and modeling techniques applied to AHSS. Finally, emerging topics such as machine learning, through-process simulation, and additive manufacturing of AHSS are discussed. The aim of this viewpoint is to identify similarities in the deformation and damage mechanisms among these various types of advanced steels and to use these observations for their further development and maturation.

show abstract

Section: Machine Learning-assisted Alloy Microstructure and Propmentioning

confidence: 79%

Current Challenges and Opportunities in Microstructure-Related Properties of Advanced High-Strength Steels

Raabe

Sun

Silva

et al. 2020

Metall Mater Trans A

137

View full text Add to dashboard Cite

show abstract

“…ML-based inverse design has attracted a great deal of attention for use in materials discovery [26][27][28][29][30] . The inverse design we were particularly interested in defines the prediction for an alloy composition, as well as the processing conditions required to accomplish a desired performance for a specific material.…”

mentioning

confidence: 99%

Dirty engineering data-driven inverse prediction machine learning model

Lee

Park

Lee

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Most data-driven machine learning (ML) approaches established in metallurgy research fields are focused on a build-up of reliable quantitative models that predict a material property from a given set of material conditions. In general, the input feature dimension (the number of material condition variables) is much higher than the output feature dimension (the number of material properties of concern). Rather than such a forward-prediction ML model, it is necessary to develop so-called inverse-design modeling, wherein required material conditions could be deduced from a set of desired material properties. Here we report a novel inverse design strategy that employs two independent approaches: a metaheuristics-assisted inverse reading of conventional forward ML models and an atypical inverse ML model based on a modified variational autoencoder. These two unprecedented approaches were successful and led to overlapped results, from which we pinpointed several novel thermo-mechanically controlled processed (TMCP) steel alloy candidates that were validated by a rule-based thermodynamic calculation tool (Thermo-Calc.). We also suggested a practical protocol to elucidate how to treat engineering data collected from industry, which is not prepared as independent and identically distributed (IID) random data.

show abstract

“…There has been considerable interest over the last few years in accelerating the process of materials design and discovery 1 . In the past decade, a new interdisciplinary research field called materials informatics, which combines data and information science with materials science, has led to an increasing number of successful materials discoveries [2][3][4][5][6][7] . Machine learning methods have played a key role in many of these studies.…”

Section: Introductionmentioning

confidence: 99%

Data augmentation in microscopic images for material data mining

Ma¹,

Wei²,

Liu³

et al. 2020

npj Comput Mater

View full text Add to dashboard Cite

Recent progress in material data mining has been driven by high-capacity models trained on large datasets. However, collecting experimental data (real data) has been extremely costly owing to the amount of human effort and expertise required. Here, we develop a novel transfer learning strategy to address problems of small or insufficient data. This strategy realizes the fusion of real and simulated data and the augmentation of training data in a data mining procedure. For a specific task of grain instance image segmentation, this strategy aims to generate synthetic data by fusing the images obtained from simulating the physical mechanism of grain formation and the “image style” information in real images. The results show that the model trained with the acquired synthetic data and only 35% of the real data can already achieve competitive segmentation performance of a model trained on all of the real data. Because the time required to perform grain simulation and to generate synthetic data are almost negligible as compared to the effort for obtaining real data, our proposed strategy is able to exploit the strong prediction power of deep learning without significantly increasing the experimental burden of training data preparation.

show abstract

A property-oriented design strategy for high performance copper alloys via machine learning

Cited by 134 publications

References 48 publications

Current Challenges and Opportunities in Microstructure-Related Properties of Advanced High-Strength Steels

Current Challenges and Opportunities in Microstructure-Related Properties of Advanced High-Strength Steels

Dirty engineering data-driven inverse prediction machine learning model

Data augmentation in microscopic images for material data mining

Contact Info

Product

Resources

About